/naive-automatic-postediting

Old code that learns some naïve postediting operators from existing postediting runs.

Primary LanguageRuby

Automatic postediting at GSoC 2018

This is a repository for Google summer of Code 2018 project.

More information about the project can be found here: http://wiki.apertium.org/wiki/Automatic_postediting_at_GSoC_2018.

Contents

  • automatic postediting: the old algorithm for extracting and applying postediting operations.
  • data: corpora, train and test sets.
  • postediting_module: postediting module testpack which can be run in Apertium pipeline
  • ru-uk_classifier: instruments for aligning Russian - Ukrainian OpenSubtitles corpora.
  • scripts: helper scripts for parsing corpora and preparing train and test data.
  • toolbox: toolbox for learning and applying postediting operations.